---
title: "EU AI Act Article 50: transparency duties from 2 August 2026"
description: "Why Article 50 reaches you, even outside the EU Article 50 of the EU AI Act (Regulation (EU) 2024/1689) is the part of the Act that touches everyday…"
url: https://optimizeall.com/learn/responsible-ai-disclosure-and-compliance/eu-ai-act-article-50-transparency
updated: 2026-10-05
---

Responsible AI, Disclosure & Compliance · Transparency law, content provenance and Gulf rules · lesson 7 of 11 · 16 min

# EU AI Act Article 50: transparency duties from 2 August 2026

## Why Article 50 reaches you, even outside the EU

Article 50 of the EU AI Act (Regulation (EU) 2024/1689) is the part of the Act that touches everyday marketing and content work. It applies from **2 August 2026**. It covers **providers** (who build or place AI systems on the market) and **deployers** (who use them in a professional capacity). The Act also reaches providers and deployers outside the EU where the **output is used in the EU**. A Lahore agency producing a realistic AI video for a Berlin client's EU campaign, or a Dubai brand running an AI chatbot for customers in Paris, can be in scope.

This lesson reflects the position as of September 2026. Always check the current consolidated text on EUR-Lex and the Commission's AI Act pages before relying on a detail.

## The four duties in plain language

| Paragraph | Who | Duty | Typical marketing example |
|---|---|---|---|
| 50(1) | Providers of AI systems that interact directly with people | Design them so people are told they are interacting with AI, unless that is obvious to a reasonably well-informed person | Website chatbot, WhatsApp AI agent, AI voice agent |
| 50(2) | Providers of systems generating synthetic audio, image, video or text | Mark outputs in a machine-readable format so they are detectable as AI-generated or manipulated, as far as technically feasible | Image, video, voice and text generators |
| 50(3) | Deployers of emotion recognition or biometric categorization systems | Inform the people exposed | In-store analytics estimating mood or age (where lawful at all) |
| 50(4) | Deployers who generate or manipulate image, audio or video that is a **deepfake** | Disclose that the content is artificially generated or manipulated | A realistic AI video of a real founder, place or event |
| 50(4) | Deployers who generate or manipulate **text published to inform the public on matters of public interest** | Disclose, unless the text has had human review or editorial control and someone holds editorial responsibility | An AI-written news-style article about an election or public health |

The information must be given **clearly and distinguishably, at the latest at the time of first interaction or exposure**, and meet accessibility requirements.

**Deepfake** in the Act means AI-generated or manipulated image, audio or video content that resembles existing persons, objects, places, entities or events and would falsely appear to a person to be authentic or truthful. Where content is evidently part of an **artistic, creative, satirical, fictional or analogous work**, the disclosure duty is lighter: you disclose the existence of generated or manipulated content in an appropriate way that does not spoil the work.

## What changed in 2026

- **Code of Practice.** In June 2026 the Commission published the final, voluntary **Code of Practice on marking and labelling of AI-generated content**. Section 1 covers machine-readable marking and detection for providers; Section 2 covers labeling of deepfakes and public-interest text for deployers, including a publicly available **EU icon** (or equivalent labels) and guidance on where and how to display it. Following the Code is not mandatory, but it is expected to be a key reference for showing compliance.
- **Digital Omnibus grace period.** The Digital Omnibus on AI, adopted in mid-2026, gives providers of generative systems placed on the market **before 2 August 2026** until **2 December 2026** to meet the 50(2) marking duty. The chatbot, deepfake and public-interest-text duties have **no** grace period: they apply from 2 August 2026.
- **Penalties.** Infringing Article 50 can attract fines of up to EUR 15 million or 3% of worldwide annual turnover, whichever is higher; for SMEs and start-ups the lower of the two applies. National market surveillance authorities enforce.

## What it means for creators, agencies and brands

1. **Chatbots and voice agents** (you are usually the deployer, but if you build and brand the bot yourself you may be the provider): open with an AI disclosure, and never design a persona that denies being AI when asked.
2. **Deepfakes in ads and content:** realistic AI video or images of real people, real places or real events need a visible label. A cloned voice of a real person is a deepfake for this purpose. Stylized, obviously illustrated work generally is not.
3. **Public-interest text:** AI-drafted articles on politics, public health or public events published to inform the public need disclosure unless a named human editor reviewed them and takes responsibility. Keep evidence of that review.
4. **Your vendors' marking:** you will rely on tool providers for 50(2) marking. Ask them how outputs are marked (metadata, watermark or both) and do not strip it in your export pipeline.

## Worked example: a Lahore agency, an EU client

A Lahore agency produces a spring campaign for a Dutch homeware brand selling across the EU.

| Asset | Article 50 analysis | Action |
|---|---|---|
| AI-generated abstract backgrounds | Not a deepfake (would not falsely appear authentic) | No label required; keep Content Credentials |
| Photorealistic AI "customer" in a real Amsterdam street | Real place shown in a scene that did not happen; would appear authentic | Label with the EU icon or "AI-generated scene"; also check consumer law if it implies a real customer |
| Founder's cloned voice-over, with written consent | Resembles an existing person: deepfake | Spoken or on-screen "AI-generated voice"; keep consent record |
| Website AI assistant | 50(1) chatbot disclosure (the vendor designs it; the agency configures the opening message) | First message states it is an AI assistant, with a route to a human |

## Hands-on: an Article 50 screening checklist and snippets

Add this to your campaign brief template and run it for every EU-facing asset.

```text
ARTICLE 50 SCREEN (one row per asset)
1. Will the output be used or seen in the EU?                  no -> stop here (other rules may still apply)
2. Is it an AI system people interact with (chat/voice)?       yes -> 50(1): disclose at first interaction
3. Is it image, audio or video that resembles a real person,
   object, place, entity or event AND would appear authentic?  yes -> 50(4) deepfake: label at first exposure
4. Is it evidently artistic, satirical or fictional?           yes -> lighter label that does not spoil the work
5. Is it text published to inform the public on a matter of
   public interest?                                            yes -> label, unless human editorial review + named editor
6. Did our tool vendor mark the file (metadata/watermark)?     unknown -> ask vendor; do not strip on export
7. Evidence saved (screenshot of label, editor sign-off)?      no -> fix before publishing
```

```text
Chatbot first message:
"Hi, I'm Lumi, the AI assistant for Casa Verde. I can help with orders and sizes.
Type HUMAN at any time to reach our team."

Video end card or lower third (deepfake):
"This video contains AI-generated imagery and voice."

Audio-only ad (spoken at the start):
"This ad uses an AI-generated voice."

Public-interest article footer (when NOT human-reviewed):
"This article was generated with AI."
```

## Measuring compliance

- 100% of EU-facing chatbots and voice agents open with an AI disclosure (test monthly, in every language you support).
- Every EU-facing asset has an Article 50 screen row in the asset register.
- Vendor list records how each generative tool marks outputs and the date you last checked.

## Pitfalls

- Assuming "we are not in the EU" ends the analysis when the output is used in the EU.
- Believing the December 2026 grace period covers deepfake labels. It only concerns providers' machine-readable marking.
- Hiding the disclosure in terms and conditions instead of showing it at first exposure.
- Relying only on a platform's automatic label to meet your own deployer duty.

## Video lecture: EU AI Act Article 50: transparency duties from 2 August 2026

Lecture coming soon · 13 chapters · about 9 minutes. Read the full transcript below.

1. EU AI Act Article 50
2. Why it matters
3. Two roles
4. The four duties
5. What's a deepfake?
6. 2026 changes
7. Example 1: Riyadh retailer's chatbot
8. Example 2 (illustrative): Lahore agency, Dutch client
9. Watch me do it: 30-second ad
10. Common mistakes
11. Ask your vendors
12. Measure it
13. Recap + try this now

## Lecture transcript

### EU AI Act Article 50

Second of August, twenty twenty-six. That's the day a set of transparency rules in the EU AI Act started to apply, and they reach far beyond Europe. If you run a chatbot that talks to customers in Paris, or you make a realistic AI video for a client selling in Berlin, they can apply to you, even if your office is in Lahore, Dubai or Manchester. In this lecture you'll learn the four Article 50 duties, who carries each one, what changed in twenty twenty-six, and a screening checklist you can run on every asset.

### Why it matters

Why does it matter? Because the Act applies to providers and deployers outside the EU when the output is used in the EU. Fines for breaking Article 50 can reach fifteen million euros or three percent of worldwide turnover, whichever is higher, with the lower figure for small businesses. But the bigger practical risk for most of you is simpler. EU clients will now ask you to prove you've done it. Agencies that can answer quickly, with evidence, win the work.

### Two roles

Here's the mental model. Think of the Act as a restaurant with two roles. The kitchen, the provider, builds the dish and must put an allergen code on the packaging that machines can read. The waiter, the deployer, serves it to the customer and must say out loud when something could be mistaken for something it isn't. Both have duties. Knowing whether you're the kitchen or the waiter for each tool is the first step. Most marketers are deployers. But if you build and brand your own chatbot, you may be a provider too.

### The four duties

Now the four duties. Paragraph one: providers of AI systems that interact with people must design them so people know they're talking to AI, unless that's obvious. Paragraph two: providers of generative systems must mark outputs in a machine-readable way, as far as technically feasible. Paragraph three: deployers of emotion recognition or biometric categorization must inform people. And paragraph four: deployers must disclose deepfakes, and must disclose AI-generated text published to inform the public on matters of public interest, unless a human editor reviewed it and takes responsibility. All disclosures must be clear, and given at the latest at first interaction or exposure.

### What's a deepfake?

What counts as a deepfake? The Act's definition: AI-generated or manipulated image, audio or video that resembles existing persons, objects, places, entities or events and would falsely appear to be authentic. So a stylized illustration of a coffee cup? Not a deepfake. A photorealistic video of a real founder saying new words, or a real Amsterdam street in a scene that never happened? Deepfake. A cloned voice of a real person? Deepfake. And if the content is evidently artistic, satirical or fictional, the duty is lighter: you disclose in a way that doesn't spoil the work, like a credit at the end.

### 2026 changes

What changed this year? Two things. First, in June twenty twenty-six the Commission published the final Code of Practice on marking and labelling AI-generated content. It's voluntary, but it's the practical reference. It covers machine-readable marking for providers, and labeling for deployers, including a common EU icon you can use. Second, the Digital Omnibus on AI gave providers of generative systems already on the market before August second a grace period to December second, twenty twenty-six, for machine-readable marking only. Chatbot disclosure, deepfake labels and public-interest text? No grace period. They apply now.

### Example 1: Riyadh retailer's chatbot

First example, simple. A Riyadh furniture retailer launches an AI chat assistant on its website, and it sells to customers in Germany and France. The assistant's first message says: Hi, I'm the AI assistant for the store. I can help with sizes and delivery. Type human any time to reach our team. They test it monthly in Arabic, English, German and French. That's paragraph one handled, whether they count as provider or deployer, because the disclosure is designed in from the first message.

### Example 2 (illustrative): Lahore agency, Dutch client

Second example, a realistic business scenario with illustrative details. A Lahore agency makes a spring campaign for a Dutch homeware brand. Four assets. AI abstract backgrounds: not deepfakes, no label, keep the metadata. A photorealistic AI customer in a real Amsterdam street: deepfake, so it gets the EU icon and an AI-generated scene line, and they check it doesn't imply a real customer testimonial. The founder's cloned voice-over, with written consent: deepfake, so a spoken AI-generated voice line. And the website assistant: an opening disclosure. Illustratively, when the client's legal team asked for the agency's Article 50 evidence, the agency sent the asset register and screenshots the same afternoon.

### Watch me do it: 30-second ad

Watch me do it. I'm screening a thirty-second video ad for an EU skincare launch. Question one: will it be seen in the EU? Yes. Question two: is it a chatbot or voice agent? No. Question three: does anything resemble a real person, place or event and look authentic? The presenter is an AI-generated woman who doesn't exist, but the lab scene shows the brand's real factory, altered to look bigger. That's a real place shown in a way that didn't happen: deepfake. Question four: evidently fictional? No. Question five: public-interest text? No. Question six: did our tool vendor mark the file? I check the vendor sheet: yes, metadata. Question seven: I add an end card, this video contains AI-generated imagery, screenshot it, and log the row.

### Common mistakes

Common mistakes. Thinking we're not in the EU ends the conversation, when the output is used there. Believing the December grace period covers deepfake labels; it only covers providers' machine marking. Hiding disclosures in terms and conditions instead of at first exposure. Relying on a platform's automatic label to meet your own duty. Forgetting audio: in an audio-only ad, say it. And stripping metadata in your export settings, which undermines the provider's marking you're relying on.

### Ask your vendors

Because you rely on your tool vendors for machine-readable marking, ask them four questions, in writing. One: how do you mark generated outputs, with metadata, a watermark, or both? Two: is marking on by default for our plan, or do we have to switch it on? Three: which of your export or download options remove the marking? And four: do you follow the EU Code of Practice on marking and labelling, and if not, how do you meet Article fifty, paragraph two? Put the answers in your vendor list with the date. If a vendor can't answer, that's a signal to consider another tool for EU-facing work.

### Measure it

How do you prove it's working? Three measures. Every EU-facing chatbot and voice agent opens with an AI disclosure, tested monthly in every language you support. Every EU-facing asset has an Article 50 screen row in your asset register. And your vendor list records how each generative tool marks outputs, metadata, watermark or both, and when you last checked. With those three, you can answer a client or a regulator in minutes, not weeks.

### Recap + try this now

Recap. Article 50 applies from the second of August, twenty twenty-six, and follows your output into the EU. Providers disclose chatbots and mark generated content; deployers label deepfakes and unreviewed public-interest text. The grace period to December is only for providers' machine marking. Use the Code of Practice and the EU icon as your reference. Try this now: add the seven-question screen to your campaign brief template and run it on three assets you're working on this week. Next, we open up Content Credentials and watermarks.

## Key takeaways

- Article 50 applies from 2 August 2026 and reaches non-EU businesses whose AI output is used in the EU.
- Providers must disclose chatbots and machine-mark synthetic outputs; deployers must label deepfakes and unreviewed public-interest text.
- The Digital Omnibus grace period to 2 December 2026 covers only 50(2) marking for systems already on the market.
- The voluntary Code of Practice and EU icon are the practical reference for how and where to label.

## Try it

Run the seven-question Article 50 screen on three assets from a current or recent EU-facing campaign and record the result and evidence in your asset register.

- [Previous: Advertising and influencer disclosure when AI is involved](https://optimizeall.com/learn/responsible-ai-disclosure-and-compliance/advertising-disclosure-and-ai)
- [Next: Content Credentials (C2PA), watermarks and provenance](https://optimizeall.com/learn/responsible-ai-disclosure-and-compliance/c2pa-content-credentials-and-watermarks)
- [All lessons of Responsible AI, Disclosure & Compliance](https://optimizeall.com/learn/responsible-ai-disclosure-and-compliance)